Quantifying vulnerability of Antarctic ice shelves to hydrofracture using microwave scattering properties

K. E. Alley1, T. A. Scambos, J. Z. Miller, D. Long, M. MacFerrin

1. College of Wooster, Ohio (current address), University of Colorado Boulder

Recent ice shelf disintegrations on the Antarctic Peninsula and subsequent increases in ice sheet mass loss have highlighted the importance of tracking ice shelf stability with respect to surface melt ponding and hydrofracture. In this study, we use active microwave scatterometry to assess melt season duration and the relative concentration of refrozen ice lenses in Antarctic ice shelf firn. We demonstrate a physical relationship between melt days and backscatter using scatterometry and field data from Greenland, and apply the observed relationship to derive and map a vulnerability index across Antarctica’s ice shelves. The index reveals that some remaining Antarctic Peninsula ice shelves have already reached a firn state that is vulnerable to hydrofracture. We also show that the progression of an ice shelf toward vulnerability is affected by many factors, such as surface mass balance and ice shelf geometry.

Surface Meltwater Ponding and Drainage Causes Ice-Shelf Flexure

Alison F. Banwell1, Ian C. Willis1, Grant J. Macdonald2, Becky Goodsell3 & Douglas R. MacAyeal2

1. Scott Polar Research Institute, University of Cambridge, Lensfield Road, Cambridge, UK.
2. Department of the Geophysical Sciences, University of Chicago, Chicago, IL, USA.
3. Antarctica New Zealand, Christchurch, NZ.

A trigger of ice-shelf disintegration is thought to be surface-stress variations associated with surface melt and ponding, causing weakness and fracture. For example, various modelling and remote-sensing studies have attributed the widespread break-up of the Larsen B Ice Shelf in 2002 to the drainage of >2000 lakes in the days preceding the rapid break-up. Until now, however, no study has provided field-based data to prove that ice-shelf flexure (and potential fracture) is indeed a direct result of surface meltwater ponding and drainage. Here, we present field data from the austral summer of 2016/2017 to show that the filling and draining of four surface lakes on the McMurdo Ice Shelf in Antarctica has a significant and simultaneous effect on vertical ice-shelf deflection and flexure. Water-depth measurements from pressure sensors reveal that lakes fill to > 2 m in depth and subsequently drain over multiple week timescales. In response, the magnitude of the vertical ice-shelf deflection, measured by 12 differential GPS receivers deployed over three months, varies according to the distance from the maximum load change. As a result, the magnitudes of the flexure stresses are high enough to initiate fracture production. By enabling models of ice-shelf flexure to be constrained, our results will enable the vulnerability and break-up potential of other ice-shelves to be constrained.

Ice-shelf stability and the importance of ice shelves to Antarctic Peninsula ice-sheet model forecasts.

Nicholas E. Barrand

School of Geography, Earth and Environmental Sciences, University of Birmingham, UK.

The southward progression of episodes of ice-shelf retreat and collapse in the Antarctic Peninsula suggests a common climatic cause and raises questions about the future stability of the remaining large shelves. Here I summarize recent attempts to incorporate ice-shelf retreat and collapse into forecasts of the volume evolution of the Antarctic Peninsula ice sheet. Surface mass balance anomalies predict ice-sheet thickening due to increased accumulation, yet these gains are dwarfed by losses of ice in response to imposed ice-shelf removal and grounding-line retreat at individual large drainage basins. Simulations of all 199 ice-shelf terminating drainage basins with shelf presence/absence determined by thermal viability limits suggest contributions of up to 10 mm to sea level in the next two centuries, the majority of these coming from basins feeding George VI Ice Shelf. An improved parameterisation of grounding-line retreat and inclusion of tidewater glaciers in the northern AP results in forecasts ranging between 11 and 32 mm sea-level equivalent to 2300, depending on the emission scenario used. The application of ice-sheet models of varying complexity (including higher-order models) to Larsen C and George VI embayment basins demonstrates the varying and relative importance to sea-level of the large AP ice shelves considered to be most at risk of collapse. Despite its large size, collapse of Larsen C does not result in large additional discharge from its tributary glaciers in any of our model scenarios. In contrast, inland ice response to collapse of George VI Ice Shelf may add up to 8 mm to global sea levels alone by 2100 and 22 mm by 2300, due in part to activation of the marine ice sheet instability mechanism. These findings highlight the crucial role of ice shelves in stabilizing the APIS and the importance of accurate estimates of collapse timing in APIS ice-sheet model forecasts.

Ross Ice Shelf front morphology from high-resolution airborne laser altimetry

Maya K. Becker1 Helen A. Fricker1, Laurence Padman2, Robin E. Bell3, Till J. W. Wagner4, Cyrille Mosbeux1, Caitlin Dieck Locke3, and the ROSETTA-Ice Team

1. Scripps Institution of Oceanography, University of California, San Diego, La Jolla, CA;

2. Earth & Space Research, Corvallis, OR;

3. Lamont-Doherty Earth Observatory, Columbia University, Palisades, NY;

4. University of North Carolina Wilmington, Wilmington, NC

The Ross Ocean and ice Shelf Environment and Tectonic setting Through Aerogeophysical surveys and modeling (ROSETTA-Ice) project combines airborne glaciological, geological,and oceanographic observations to enhance understanding of the history and dynamics of Ross Ice Shelf (RIS). We focus on data from the scanning laser altimeter, which provides raw elevation measurements at a density of 1-4 points/m2 across a ∼750 m wide swath for most of the ROSETTA-Ice grid lines. With this high resolution, we are able to image fine-scale features such as the shape of the RIS ice front. Several front-crossing profiles show a trough inboard of a raised edge, much like the “rampart-moat” morphology that has been previously observed on and modeled for icebergs. This shape arises from preferential melting of the ice front above the waterline, which in turn leads to the development of a submerged bench of ice whose buoyancy pushes the top edge upwards and, eventually, results in a calving event. We also observe this feature on RIS in data from the Ice, Cloud and land Elevation Satellite (ICESat). The ICESat data are useful because they provide a time series of the evolution of the rampart-moat profile from a smoother berm-like profile and the subsequent calving. In addition, we present preliminary results from two modeling exercises. The first compares the ROSETTA-Ice observations with those theoretically predicted by an elastic beam model that has recently been applied to icebergs and glacier termini. The second employs the Elmer/Ice finite-element model to examine the effects of the inclusion of an underwater bench on the shape and stress regime of the RIS ice front. Missing from these models, however, is the influence of surface meltwater, which may collect in the moats. Moreover, meltwater-accelerated fracturing could fundamentally alter the calving style of ice-shelf fronts that exhibit the rampart-moat morphology.

Are surface meltwater features in Greenland suitable analogs for the Antarctic?

Alexandra L. Boghosian1 and Laura A. Stevens1

1. Lamont-Dohery Earth Observatory, Palisades, New York, USA.

Antarctic ice shelves are projected to experience increased surface meltwater production over the coming century. However, the impact of surface meltwater on ice shelf stability and ice sheet surface mass balance is unknown. Additionally, it is unknown which dominant surface meltwater features will develop. The Greenland Ice Sheet ablation zone at present hosts meltwater features that may give us a preview of the future surface hydrology of Antarctic ice shelves. In this guided group discussion, we will examine what, if any, supra- and en-glacial meltwater features observed on the Greenland Ice Sheet may serve as suitable analogs for understanding surface hydrology of the present and future Antarctic Ice Sheet. Both the grounded and floating portions of these ice sheets will be considered. We will focus the discussion around four questions. First, can equivalent meltwater features develop over bare ice in the Greenland Ice Sheet ablation zone and blue ice and/or densified firn in Antarctica? Second, is water movement into and through the firn in Greenland analogous to water movement into and through the firn in Antarctica? Third, will surface hydrology on Antarctic ice shelves eventually resemble the surface hydrology of the two largest remaining ice shelves in Greenland? Finally, could you develop a supraglacial and englacial hydrologic system over grounded ice in Antarctica analogous to the present hydrologic system in the Greenland ablation zone? Drawing on the expertise of workshop attendees, this discussion will illuminate priorities for both Greenlandic and Antarctic surface hydrology research efforts by identifying how we can best use knowledge of the established surface hydrologic network in Greenland to better understand and predict present and future Antarctic surface hydrology.

Can we use the past to better predict future Antarctic surface melt?

Sarah B. Das1, Luke D. Trusel2, Laura A. Stevens3

1. Department of Geology and Geophysics, Woods Hole Oceanographic Institution

2. Department of Geology, Rowan University

3. Lamont-Doherty Earth Observatory, Columbia University

Surface melting is intricately linked to many important dynamical processes that contribute to ice mass loss.  Yet our understanding of ice sheet and ice shelf surface melting is still extremely limited.  Beyond the relatively short satellite era, records of past melt occurrence and intensity are sorely lacking, and depend almost entirely on climate model reconstructions. This limits our ability to understand how climate influences surface melt variability both today and in the future.  This is a particularly acute problem for Antarctica, where surface melting is highly variable from year to year.  Additionally, outside of the Antarctic Peninsula and northerly ice shelves, surface melting still constitutes a ‘rare event’, making it further challenging to interpret the significance of satellite-era observations.  Fortunately, surface melting fundamentally alters the physical nature of near surface snow, leaving behind a physical ‘fingerprint’ of melting in the firn.  This melt-layer stratigraphy, which can be derived from ice-cores or borehole measurements, provides a unique and invaluable record of decadal- to millennial-scale Antarctic melt history.  Despite these measurements’ importance, however, few such records exist, and they are geographically limited.  We suggest advances in two areas will greatly improve our use of stratigraphic melt records to better understand modern and project future Antarctic surface melt: (1) longer and more optimally sited records; and (2) improved quantification of past melt intensity from these records.  We describe some ongoing efforts, as well as future goals, towards both.  In particular, we first assess what is known about Antarctic surface melt variability over the last millennium. Then, we present initial efforts to use modern stratigraphy observations of recent melting, including the Ross Ice Shelf 2016 event, in concert with remotely-sensed and modeled metrics quantifying surface melt, to calibrate these stratigraphic changes.

The impact of föhn winds on the Larsen C ice shelf: meltwater patterns and infiltration as simulated by the MAR regional climate model (1982 - 2017)

R.T. Datta1, Tedesco, M.2,3, Agosta, C.4, Fettweis, X.4, Kuipers Munneke, P.5, van den Broeke, M.5

1. The Graduate Center, City University of New York, NY 10016, USA

2. Lamont-Doherty Earth Observatory of Columbia University, Palisades, New York, NY, 10964, USA

3. NASA Goddard Institute of Space Studies, New York, NY, 10027, USA

4. Department of Geography, Université de Liège, Liège, Belgium

5. Institute for Marine and Atmospheric Research, Utrecht University, Utrecht, The Netherlands

The process of hydrofracture initiated by enhanced meltwater production has been implicated in ice shelf collapse on the Antarctic Peninsula, most dramatically the Larsen A (1995) and Larsen B (2002) ice shelves. Surface melt on the remaining Larsen C ice shelf can result either from regional warming or from föhn flow produced by sporadic westerly winds. Surface melt is therefore sensitive to long-term atmospheric trends, including a potential warming trend before 1998 followed by a cooling trend, as well as changes in atmospheric circulation patterns affecting the speed and direction of westerly winds (and therefore föhn flow).  Additionally, the degree to which meltwater at the surface can percolate into the snowpack is affected by firn density, which is strongly influenced by changes in the seasonal cycle for melt.

We discuss spatial patterns of meltwater production in the northeast of the Antarctic Peninsula (including the Larsen C ice shelf) as modeled by the Modèle Atmosphérique Régionale (MAR), at a 10km resolution between 1978 and 2017. The timeseries associated with these patterns are used to identify interannual trends in the frequency of primary melt patterns. These trends are used to quantify interannual trends in (a) föhn-induced melt (b) melt associated with regional temperature trends and (c) meltwater percolation depth. We will examine the potential implications of these trends for surface hydrology on the northern Larsen C ice shelf.

Changing surface hydrology, strain rates and fracture distribution across Nivlisen and Riiser-Larsen Ice Shelves, Queen Maud Land, Antarctica

Becky Dell1, Ian Willis1 , Alison Banwell1 , Neil Arnold1 , Hamish Pritchard2

1. Scott Polar Research Institute, Cambridge University, Lensfield Road, Cambridge, CB2 1ER

2. British Antarctic Survey, High Cross, Madingley Road, Cambridge, CB3 0ET

The collapse of four major Antarctic Peninsula ice shelves since the 1950s, most notably the catastrophic collapse of the Larsen B Ice Shelf in 2002, has highlighted the need for a greater understanding of the drivers of ice shelf instability under current climate warming trends. In particular, surface and basal melting, firn densification, ponding, vertical hydrofracturing, horizontal fracture propagation, and ice-shelf edge retreat have all been identified as factors that may have contributed to past ice shelf collapse events. In order to further investigate the potential precursors to ice-shelf instability, we are making a series of remotely sensed observations of the Nivlisen and Riiser-Larsen Ice Shelves, using a combination of Optical Imagery from Sentinel-2 and Landsat, and Synthetic Aperture Radar (SAR) Imagery from Sentinel-1 and ERS 1 and 2. Using these data, we are developing semiautomated methods for tracking spatial and temporal changes in: i) surface and subsurface meltwater ponding; ii) surface velocity and strain rate patterns; and iii) the density and geometry of surface fracture patterns. The analysis and synthesis of these data should allow us to develop a vulnerability index to show how close an ice shelf is to potential instability and collapse, and how this vulnerability may change under future climate warming scenarios.

Mario Garnsworthy

Antarctic ice sheet collapse would drastically change our planet, causing massive disruption to civilization. But to the public, ice shelves are remote, foreign, and easily pushed from their consciousness. Information may be imbibed via alarmist headlines, or worse, via sources whose primary mission is to misinform regarding climate change-related issues.

So, how do polar scientists best engage the public with accurate information?

Effective science communication is one way. Science communication and outreach are necessary but take time from research, teaching, and other job necessities. It can also be challenging for scientists to convey information in a manner that will interest or be understood by the person on the street.

Art is another. Polar regions are places of spellbinding natural beauty, and since the early days of polar exploration, artists have been instrumental in sharing them with the public. Narrative non-fiction is yet another, and it’s human nature to prefer information delivered via story. (E.g., though the Antarctic Writers and Artists program, the NSF realizes the vital role creatives play in communicating and promoting science.) Artists and writers are in a unique position to engage and educate the public via means that inspire, inform, and intrigue, rather than those that alarm, confuse, and repel.

I combine these three skill-sets—science communication, art/illustration, and writing—along with vast social media experience. I am the science communicator and outreach ambassador for our Antarctic SNOWBIRDS Transect research cruise. I blog at POLAR BIRD, my mission to inform and involve the public with succinct, easily understood information about the nature and state of our polar ice—my greatest driving passion. I am not a scientist but a published author/illustrator, and I hope to attend this workshop to build on my knowledge of ice shelves, to create relationships with polar scientists, and to explore future collaboration.

Marlo Garnsworthy www.wordybirdstudio.com wordybirdsci.com

www.snowbirdstransect.org Book Editing Associates 

The hydrology of ice shelves: processes and implications for dynamics

Arno Hammann1, Shelley MacDonell1, Francisco Fernandoy2, Gino Cassasa3 , and Rémi Valois1 

1. Centro de Estudios Avanzados en Zonas Áridas (CEAZA), La Serena, Chile

2. Universidad Andrés Bello, Viña del Mar, Chile

3. Universidad de Magallanes, Punta Arenas, Chile

We propose to study, in situ and through numerical modelling, the processes related to meltwater production and movement on and within ice shelves and how they affect ice shelf dynamics. Meltwater has been observed in isolated studies to run off or pool on ice shelves, but also to refreeze within the firn; but more observations are needed to gain a representative understanding of the role of liquid water on ice shelves. This is particularly important since hydrofracturing and shelf flexure under the weight of liquid water have been identified as leading causes for ice shelf disintegration.

Three fundamental goals of our study are: quantification of meltwater generation and refreezing, the study of the water’s movement across and through an ice shelf, and linking the observed structure of the hydrological system with ice shelf dynamics. Our currently envisaged study location is the Larsen C Ice Shelf, on which both the occurrence of surface melt pools and of subsurface refreezing have been observed and which straddles the -9°C annual mean air temperature isotherm considered the limit of ice shelf stability.

Meltwater generation will be mapped and quantified by combining the results from firn core analyses and meteorological observations with spatially distributed energy balance modelling. Meltwater movement and refreezing will be studied using satellite imagery, ground and/or areal radar surveys, electromagnetic hydrogeological methods, borehole measurements, firn/snow cores and stream gauging. Finally, we will relate the spatial characteristics of the hydrological system to ice shelf dynamics by means of forward and inverse numerical modelling approaches. In the forward approach, ice mechanical properties measured in the field will be used to characterize the parameters of a numerical model of ice dynamics, while in the inverse approach, satellite-derived flow velocity fields and GNSS measurements constrain the diagnosis of spatial inhomogeneities of ice properties.

Dominant drivers of surface melt on an Antarctic ice shelf

C.L. Jakobs1, C.H. Reijmer1, M.R. van den Broeke1, G. K ̈onig-Langlo2

1 Institute for Marine and Atmospheric Research Utrecht (IMAU), Utrecht University, The Netherlands.

2 Alfred-Wegener-Institut Helmholtz-Zentrum f ̈ur Polar- und Meeresforschung, Bremerhaven, Germany.

Thoroughly analyzing the Surface Energy Balance of snow and ice surfaces enables us to better understand the process of surface melt. Furthermore, investigating the individual energy balance components provides insight in their separate contributions to surface melt. We use 25 years of meteorological data from an Antarctic ice shelf to force an Energy Balance model in order to calculate the amount of surface melt. The measurements are taken from Neumayer station, an AWI- operated scientific research station on the Ekstr ̈om ice shelf in Dronning Maud Land, Antarctica. Turbulent fluxes for sensible and latent heat are calculated using the bulk method, subsurface processes are incorporated with a snow model that includes a simple tipping bucket scheme for water percolation. Furthermore, the effect of uncertainties in both measurements and modeling parameters such as roughness length and fresh snow density are investigated by multiple random simulations. Spatial variability of the melt season is assessed by including meteorological data from two additional automatic weather stations on nearby ice shelves.

We find that the melt season at Neumayer is mostly concentrated around December and January, with melt occurring on approximately 20 days on average. The effects of measurement uncertainties are very small, leading to an uncertainty of 1% in the modeled amount of surface melt. The uncertainties in modeling parameters affect the outcome much more, leading to a spread in modeled surface melt of 10–15%. By comparing the individual energy balance components in the years with most and least melt, we gain insight in their separate contributions. We find that the energy balance is mainly driven by radiative components, in particular the shortwave radiation. Including the results from the additional weather stations, we find the same relationship between the number of melt days and the amount of surface melt.

Present and Past Melt at the Head of Law Glacier, Near the Edge of the EAIS Plateau, 84°S

Mike Kaplan1, Kathy Licht, Gisela Winckler, Joerg Schaefer, Joseph Graly, Jennifer Lamp

1. Lamont-Doherty Earth Observatory of Columbia University, Palisades, New York, NY, 10964, USA

A blue ice area and accompanying moraine exist around Mt. Achernar, ~84.2°S, ~162°E, where plateau ice converges into Law Glacier. At this site, we show evidence of melting at >1800 meters, in the present and past. In January, 2011, we observed a small surface stream flowing near the moraine/ice boundary; other signs of melt were apparent but were not observed. In the 2015/16 season, we observed ubiquitous signs of melt. Melting occurs around sediment on the ice surface and on the blue ice moraine near its boundary with Law Glacier. Farther inside the blue-ice moraine, away from Law Glacier (e.g., >100 m), where till thickness increases (>10 cm) and blankets the underlying ice, evidence of melt was not observed.

Mt. Achernar and the central Transantarctic Mountains are ultimately responsible for the formation of blue ice moraine sediments at the locality. However, in this setting observed melt is not (only) around exposed rock, but related with till/sediment that now sits on the surface, even several kilometers away from headwalls. We also infer that the melt has led to areal surface lowering, which we observe in 10Be ages on clasts emerged on the surface recently in the last few decades.

Past melt is evident in the form of closely spaced pond-like features in the blue ice moraine. These apparent ponds often interconnect with seeming flow towards Law Glacier. In 2015/16 we found melt around the pond margin. We assume the ponds grow and interconnect (i) by melt around the borders in climates such as today and (ii) in the past, with warmer summers and more pervasive melting than we observed. Questions remain on how often and how much of the pond surface ice melts, and whether there are extensive drainage systems in and out of the features.

In-situ observation of surface melt on Antarctic Peninsula ice shelves compared to RACMO2

Peter Kuipers Munneke1 , Paul Smeets1 , Melchior van Wessem1 , Michiel van den Broeke1 , Adrian Luckman2 , Bryn Hubbard3

1. Institute for Marine and Atmospheric research Utrecht (IMAU), Utrecht University,

2. Department of Geography, Swansea University,

3. Department of Geography and Earth Sciences, Aberystwyth University

Ice-shelf collapse by surface melt-induced hydrofracturing is implied in future scenarios of rapid Antarctic ice discharge and sea-level rise. However, meltwater will fill and deepen crevasses only after degradation of the firn layer, which otherwise would act as a sponge. The state and evolution of the firn is in turn dependent on accumulation, and the meltwater flux itself. An accurate modelling of the present-day and future surface climate of ice shelves is therefore crucial for predicting future ice-shelf collapse. Determining in-situ melt fluxes is challenging, but can be achieved with reasonable accuracy using an energy balance model forced with quality-controlled automatic weather station data. Here, we present in-situ melt fluxes for 3 ice-shelf stations in the Antarctic Peninsula. At one of these locations (Cabinet Inlet on Larsen C ice shelf), the melt climate is strongly dependent on the occurrence of föhn winds descending from the Antarctic Peninsula mountains. Ponding of meltwater in Cabinet Inlet modifies temperature and structure of the ice downstream. We compare the latest RACMO2-simulation of Antarctic climate (version 2.3p2) with observed melt fluxes and conclude that the agreement is good, but with a small low bias in RACMO2. RACMO2 simulates an annual surface melt volume of 112 Gt, of which 41 Gt in the Antarctic Peninsula. These numbers can provide guidance to further development of models of ice-shelf collapse.

Polar surface temperature bias reduction achieved by including realistic longwave surface emissivity in the Community Earth System Model

Chaincy Kuo1*, Daniel R. Feldman1, Xianglei Huang2, Mark Flanner2, Ping Yang3, Xiuhong Chen2

1. Lawrence Berkeley National Laboratory, Berkeley, California

2. University of Michigan, Ann Arbor, Michigan

3. Texas A&M, College Station, Texas

Frozen and unfrozen surfaces exhibit different longwave surface emissivities with different spectral characteristics (Feldman et al. [2014]; Huang et al. [2016]), and outgoing longwave radiation and cooling rates are reduced for unfrozen scenes relative to frozen ones (Cheng et al, [2016]).  This discrepancy can affect both cryospheric change processes and ability of Earth System models to predict climate.  For example, the modeling of surface melt pond formation can be biased by several W/m2 by assuming perfect infrared emission, and this will have a cascading impact on ice sheet surface melt and hydrofracturing. On a global scale, physically-realistic modeling of spectrally-resolved surface emissivity is essential to reducing high-latitude model biases: recent work by Kuo et al [2017] explored how the Community Earth System Model (CESM) responds to the inclusion of realistic surface emissivity throughout the coupled model components. It is shown that despite a top-of-atmosphere surface emissivity feedback amplitude that is, at most, a few percent of the surface albedo feedback amplitude, the inclusion of realistic, harmonized longwave, spectrally-resolved emissivity information in CESM1.2.2 reduces wintertime Arctic surface temperature biases from −7.2 ± 0.9 K to −1.1 ± 1.2 K, relative to observations. The bias reduction is most pronounced in the Arctic Ocean, a region for which Coupled Model Intercomparison Project version 5 (CMIP5) models (Taylor et al. [2012]) exhibits the largest mean wintertime cold bias (Flato et al. [2013]), suggesting that persistent polar temperature biases can be lessened by including this physically-based process across model components. Furthermore, the new model, CESM-𝜀(n), produces an Arctic September sea-ice decline of -5.9±1.2%/decade and outperforms the 10 member CESM-LME (-2.5±0.4%/decade) and the IPCC AR4 model mean (-2.6±0.2%/decade) September sea-ice decline, with respect to the observational finding over 1953-2006 of -7.8±0.6%/decade. These results highlight the importance of infrared physics for resolving persistent CESM biases and the approach taken here can be readily adapted to other Earth System Models.

Spatial Heterogeneity of Bed Processes in Supraglacial Streams

Sasha Leidman,

Rutgers University.

Supraglacial streams are pervasive on ice sheets and disproportionately affect albedo (Ryan et al. 2016, Kingslake et al. 2017). Future warming may cause supraglacial streams to expand contributing to a positive feedback loop for melting. Supraglacial streams may also cause seasonal speedups of the ice sheet and worsen hydrofracturing exacerbating calving (Palmer et al. 2011). Understanding the flow dynamics of supraglacial streams is therefore vital for predicting Antarctic future contributions to sea level rise. Such predictions though require knowledge of how the hydrologic parameters that dictate routing vary along the flow path. During the summer of 2017, high resolution GPS measurements of the stream bed, stream velocity measurements, and sediment samples were collected on a supraglacial stream in southwest Greenland. These measurements allow for a detailed investigation of the spatial variability of bed incision and relate it to changes in bed friction, radiative melting, and sediment distribution. The evidence points to a potential velocity threshold in supraglacial streams in which a sediment heavy, radiative melting dominated system switches to a smoothed bed, frictional melting dominated regime. This complicates the idea of a steady decrease in albedo with higher flows as increased meltwater supply might flush out dark sediment-laden stream beds. These stream dynamics will likely play an important role in dictating future albedo changes on the Antarctic ice sheet.

Exploring the fate of meltwater on Antarctic ice shelves: the need of an interdisciplinary approach

Jan Lenaerts

University of Colorado

The recent discovery of surface and sub-surface meltwater pockets on an Antarctic ice shelf, where the climate has traditionally been assumed too cold for these processes to occur, indicates that many Antarctic ice shelves are vulnerable to hydrofracturing. However, it is still poorly understood how these meltwater pockets are generated, how they evolve in time, and what their impact on the ice shelf firn is. I propose to use a novel combination of firn and climate models and observations from weather stations and satellites to tackle these problems. We will assess what specific climate circumstances are required to generate and retain the liquid water, and track the extent, volume, and depth of the meltwater pockets as they flow from the ice sheet to the ocean. Finally, we will analyze the effect of the meltwater on the surrounding ice shelf, to ultimately assess the overall potential for hydrofracturing, and how that will change in a future warming climate. This highly multi-disciplinary work will require the formation of a team of experts in climatology, snow modeling, hydrology, and glaciology.

The importance of the melt-albedo feedbacks for Antarctic ice shelf melt

Stef Lhermitte1, Jan Lenaerts2

1. Delft University of Technology, Netherlands

2. University of Colorado, USA

Surface melt and subsequent firn air depletion is considered an important precursor for disintegration of Antarctic ice shelves, causing grounded glaciers to accelerate and sea level to rise. Recent studies have highlighted the impact of surface winds on Antarctic ice shelf melt, both on the Antarctic Peninsula and in East Antarctica. In the Antarctic Peninsula, foehn winds enhance melting near the grounding line, whereas on the East Antarctic ice shelves, meltwater-induced firn air depletion is found in the grounding zone as result of persistent katabatic winds. Both these foehn and katabatic winds regionally warm the atmosphere and induce a melt-albedo feedback that further enhances the meltwater production.

Here, we use a combination multi-source satellite imagery, snow modelling, climate model output and in-situ observations to highlight the importance of this melt-albedo feedback as a melt-enhancing factor. On the one hand we merge climate model output with snow modelling experiments to quantify the contribution of this melt-albedo feedback. On the other hand, we combine multi-source satellite imagery including optical, passive/active microwave and synthetic aperture radar (SAR) data sets to provide insight in the meltwater drainage systems, showing spatio-temporal changes in both supraglacial and englacial water throughout the melt season and during the subsequent winter.

Results of the optical and SAR meltwater detection methods illustrate the prominence of melt in the blue ice areas or areas with strong firn air depletion. Simultaneously they illustrate the shortcomings of current climate models and melt detection techniques based on scatterometer or passive microwave data to include these melt processes.

Early warnings of further disintegration of Pine Island Glacier’s ice shelves

Stef Lhermitte1, Bert Wouters2, Christopher Shuman3

1. Delft University of Technology, Netherlands

2. UMBC Joint Center for Earth Systems Technology, USA

3. Utrecht University, Netherlands

Pine lsland Glacier (PIG) is among the fastest changing outlet glaciers of West Antarctica, with strong changes in acceleration, retreat, and thinning in recent decades. If these changes continue, PIG is expected to have an even larger impact on accelerated future retreat and contribute to increasing mass losses from West Antarctica. Therefore, understanding evolving PIG behavior is crucial for improved future sea level rise estimations.

In this study we highlight some early warning signs of further weakening of PIG’s floating ice tongue and adjacent ice shelves by combining multi-source satellite data (optical, altimeter and SAR radar) and analysing derived acceleration, retreat, and thinning time series.

The early warnings of further PIG disintegration include changes in calving patterns and rifting dynamics. For example, over the last decade calving frequency has increased and the calved icebergs have disintegrated more rapidly as a result of changes in the rifting patterns likely caused by warm waters in Pine Island Bay. The rifting patterns have shifted from rifting in the shear zone near the northern ice shelf to internal rifting once this zone retreated. Satellite imagery since 2016 shows that a new rift zone is developing in the southern shear zone which may be related to changes in glacier velocity.  This rifting may be part of a positive feedback loop which could result in a further decoupling from the southern ice shelf and which may initiate further disintegration.

Wave-Induced Ice Shelf Rift Propagation

Bradley Paul Lipovsky

Department of Earth and Planetary Sciences Harvard University

Distant storms, tsunamis, and earthquakes generate waves in floating ice shelves. In several instances, seismic observations have suggested a mechanistic link between periods of elevated wave activity and ice shelf rift propagation. The detailed mechanical interpretation of observed seismograms is complicated, however, by the existence of numerous types of waves that propagate in the coupled ice--ocean--earth system. Here, I describe wave propagation in an elastic, finite-thickness, buoyantly floating ice layer above a uniform and inviscid water layer. I place particular focus on waves with wavelength greater than the ice thickness, as have recently been observed on the Ross, Pine Island, and Amery Ice Shelves. I show that mode uncoupling occurs at long period such that waves occur as either symmetric or flexural modes. I calculate the stresses associated with the seismically observed wave field on the Ross Ice Shelf. In the second part of this work, I place these stresses the context of linear elastic fracture mechanics. I show that long rifts in buoyantly floating ice shelves experience stabilization due to the inability of a thin elastic layer to effectively transmit stresses over long distances. I derive a rift tip equation of motion that shows excellent agreement with observed rift tip propagation velocities. The theory presented here paves the way to an improved depiction of ice shelf calving in predictive ice sheet models.

Investigating the drainage of supraglacial lakes on Antarctic Ice Shelves

Grant J. Macdonald1, Alison F. Banwell2, Andrew Williamson2 Ian C. Willis2, Douglas R. MacAyeal1

1.   Department of the Geophysical Sciences, University of Chicago, Chicago, IL 60637, USA

2.   Scott Polar Research Institute, University of Cambridge, Lensfield Road, Cambridge, CB2 1ER, UK

Evidence of rapid supraglacial lake drainage exists for several ice shelves around Antarctica.  Numerous studies hypothesize that rapid lake drainage events on the Larsen B Ice Shelf on the Antarctic Peninsula were the immediate cause of its catastrophic collapse in March 2002.  There is additional evidence from sediment cores of rapid lake drainage through the ice of Larsen B.  The presence of dolines also suggests that rapid lake drainage takes place on the Roi Baudoin Ice Shelf and dolines and observations suggest it occurs on the floating portion of the Langhovde Glacier in East Antarctica.  However, a comprehensive investigation of the frequency of rapid and slow lake drainage is currently lacking.

We use the Normalized Difference Water Index (NDWI) method to identify supraglacial lakes from Landsat 8 and Sentinel-2 images over the austral summers of 2014-17 on five Antarctic ice shelves (George VI, Larsen C, Nansen, Roi Baudouin and Pine Island Glacier).  The combination of Landsat 8 and Sentinel-2 allows for a high temporal resolution of coverage, comparable to that of MODIS but with a higher spatial resolution than MODIS.  The ‘Fully Automated Supraglacial lake area and volume Tracking Enhanced Resolution’ (FASTER) algorithm developed to identify rapid lake drainage events on the Greenland Ice Sheet is then used to calculate the volume of lakes and identify drainage events.  Rapid (<4 days) and slow (>4 days) drainage events are identified and investigated.  We then compare the results with temperature data and other data, such as strain rates, to identify any patterns associated with rapid, slow or non- lake drainage.

This study will provide important insight into the behaviour of meltwater on ice shelves in Antarctica.  In particular, understanding the frequency of rapid lake drainage events is essential to assess the implications of surface meltwater on ice shelves for ice-shelf stability.

The role of snowfall on the meltwater storage capacity of the firn column over Antarctic Ice Shelves

B. Medley, T.A. Neumann, C.M. Stevens, H.J. Zwally

From an ice-sheet surface mass balance (and admittedly selfish) perspective, the most important role of Antarctic snowfall is arguably through its ice mass contribution.  The surface environment over an ice shelf, however, is different from the grounded ice sheet (and is far from the simple flat, white landscape I envisioned) as there are significant small- and large-scale topographic features, bare ice zones, and liquid water that is organized into lakes, streams, and subsurface aquifers, to name a few distinctions.  Thus, snowfall (or lack thereof) over an ice shelf takes on additional responsibilities including (but not limited to) modification of the surface energy budget through the albedo change and modulation of internal accumulation (or the percolation and refreezing of surface meltwater within the snow/firn).  Here, we focus on the latter and ask the very basic question of: What role does snowfall play on meltwater retention, refreezing, and runoff over climatically-evolving Antarctic Ice Shelves?

This firn buffer is likely temporary (as in the case of Greenland) if the rate of meltwater production outpaces increases in snowfall.  But, as is not the case in Greenland, many of the ice shelves receive a large amount of snowfall (in some cases, well over 1 meter of water equivalence), and they are located adjacent to their moisture source, which could make their surface climate more susceptible to rapid and large changes.  We force the Community Firn Model (CFM) with precipitation-minus-evaporation and temperature fields from the Community Earth System Model over climatically distinct ice shelves to quantify the projected increases in the firn porosity due to enhanced snowfall.  Initial results suggest firn porosity is increasing over many of the ice shelves due to recent snowfall increases.  Next, we apply very simple melt thresholds to the CFM runs to determine the melt fluxes necessary to overwhelm the snowfall signal.

Meltwater percolation and refreezing in compacting snow

Colin R. Meyer1 and Ian Hewitt2

1. Department of Earth Sciences, University of Oregon, Eugene, OR 97405

2. Mathematical Institute, Woodstock Road, Oxford, OX2 6GG

Meltwater is produced on the surface of glaciers and ice sheets when the seasonal energy forcing warms the snow to its melting temperature. This meltwater percolates into the snow and  subsequently either runs off laterally in streams, is stored as liq-uid water, or refreezes, hence warming the subsurface through the release of latent heat. We present a continuum model for the percolation process that includes heat conduction, meltwater percolation and refreezing, as well as mechanical compaction. The model is forced by surface mass and energy balances, and the percolation process is described using Darcy’s law, allowing for both partially and fully saturated pore space. Water is allowed to runoff from the surface if the snow is fully saturated. The model outputs include the temperature, density, and water-content profiles as well as the surface runoff and water storage. We compare the propagation of one-dimensional freezing fronts that occur in the model to observations from the Greenland ice sheet. These observations also show that meltwater flows laterally, which we capture in two- dimensional simulations. Our model applies to both accumulation and ablation areas and allows for a transition between the two as the surface energy forcing varies. The largest average firn temperatures occur at intermediate values of the surface forcing when perennial water storage is predicted.

Quantifying meltwater in supraglacial lakes across Antarctica from a suite of satellite observations

Mahsa Moussavi1, Allen Pope1, Luke Trusel2,

1 National Snow and Ice Data Center (NSIDC), Univesity of Colorado, Boulder

2 Rowan University

Melting of snow and ice at the surface of the Antarctic ice sheet leads to the formation of meltwater lakes, an important precursor for ice shelf collapse and accelerated ice sheet mass loss. Owing to the nonlinear rise in surface melt rates under climatic warming, accurate prediction of Antarctic sea level contributions critically requires understanding present melting and supraglacial lake conditions. This study will quantify meltwater contained in supraglacial lakes across Antarctica by using data from a suite of Earth-observing satellites. This research specifically seeks to answer the following questions: (1) when and where do supraglacial lakes form in Antarctica? and (2) how much water is stored in Antarctic supraglacial lakes? The fundamental concept behind spaceborne bathymetry is to use depth-reflectance relationships and to isolate the effects of depth, optical properties of water column, and the bottom albedo on measured reflectance. Many studies in Greenland have used physically based and empirical passive remote sensing techniques to derive bathymetric information over supraglacial lakes using the Moderate Resolution Imaging Spectro-radiometer (MODIS), the Advanced Spaceborne Thermal Emission and reflection Radiometer (ASTER), Landsat 7, and Landsat 8. Given the smaller size and shallower nature of supraglacial lakes in Antarctica, we will use higher resolution Landsat series, complemented by Sentinel-2, and augmented by MODIS. The combination of these sensors will allow us to study supraglacial lake distribution and volume in Antarctica over 30+ years, yielding an unprecedented understanding of lake behavior in Antarctica. First, however, this presentation will show progress toward this goal by beginning with various case study areas around the perimeter of Antarctica.

Coordination for multi-centennial ice core perspectives on coastal WAIS / Marie Byrd Land climate variability

Peter D. Neff

University of Washington

The trajectory of observed ice shelf change along the coast of the West Antarctic Ice Sheet (WAIS) is a large uncertainty in sea-level rise projections. The role that surface melt will play is not well known but projected to become more significant (Trusel et al., 2015). Though it has long been known that the El Niño-Southern Oscillation (ENSO) and Southern Annular Mode significantly influence coastal WAIS climate (e.g. Renwick, 2002), it has only recently been shown how these patterns force ice-shelf dynamics (Paolo et al., 2018) and surface melt events (Abram et al., 2013; Nicolas et al., 2017). Annually-resolved, spatially-complete satellite records of ice shelf behavior and surface melt allow for calibration of ice core proxy reconstructions (e.g. marine salt wind proxies, sulfur-based surface ocean proxies, water stable-isotope temperature proxy). This provide centennial to multi-centennial perspective on ice-shelf behavior and surface melt variability and trends in this region. This is needed to evaluate the significance of the 1940s El Niño event, which may have initiated modern retreat of Pine Island Glacier (Schneider and Steig, 2008; Smith et al., 2017), and likely also forced surface melt across WAIS ice shelves (temperature anomaly of ~0.3ºC; Schneider & Steig, 2008). Existing ice core records of melt frequency are insufficient for assessing coastal WAIS, available only at Siple Dome (Siple Coast, Ross Ice Shelf; Das and Alley, 2008) and James Ross Island (northern Antarctic Peninsula, Abram et al., 2013). Investigating the influence of natural climate variability on both ice shelf ice-dynamics and surface hydrology requires that remote field studies are designed to inform both of these subjects, and also capture the longitudinal gradient of ENSO influence from the Amundsen Sea coast to the Ross Ice Shelf (greatest at Sulzberger Ice Shelf, where, for instance, significant surface melt is observed near the Phillips Mountains).

Characterizations of stream hydraulics and the hourly hydrograph in the Rio Behar, a large supraglacial river on the Greenland Ice Sheet

Lincoln Pitcher1, Laurence Smith1, Brandon Overstreet2, Matthew Cooper1, Charlie Kershner3, Asa Rennermalm4, Jonathan Ryan1, Claire Simpson1

1. Department of Geography. University of California – Los Angeles

2. Department of Geography. University of Wyoming.

3. Department of Geography and Geoinformation Science. George Mason University.

4. Department of Geography. Rutgers University.

Rivers on ice sheets, ice shelves, and glaciers connect surface climatology with ice dynamics processes and thus are integral to simulations of the response of the cryosphere to climate variability and eustatic modifications to global sea levels. However, particularly in comparison to their alluvial counterparts, relatively little is known about the genesis, evolution, hydrology, hydraulics, or impact of supraglacial rivers on glacier, ice sheet, and ice shelf processes. To that end, this work analyzes detailed in situ streamflow measurements collected in a large moulin terminating channel on the western Greenland Ice Sheet to investigate both stream hydraulics including channel incision and the effects of slush on water velocity and the hourly hydrograph. We argue that this analysis helps improve process-level understanding of supraglacial hydrology and thus is useful to enhancing knowledge of Antarctic Surface Hydrology.

Development of a Snow-Firn-Ice Surface Mass Balance Treatment for Ice Sheet Models

David Pollard(1), Robert M. DeConto (2).

1. Earth and Environmental Systems Institute, Pennsylvania State University, University Park, Pennsylvania, USA.

2. Department of Geosciences, University of Massachusetts, Amherst, Massachusetts, USA.

The treatment of surface melt, runoff, and the snow-firn-ice transition in ice-sheet models (ISMs) is becoming increasingly important, as mobile liquid on Greenland and Antarctic flanks increases due to climate warming in the next century and beyond. Simple PDD-based box models used in some ISMs crudely capture liquid storage and refreezing, but need to be extended to include vertical structure through the whole firn-ice column, as in some regional climate models (RCMs). This is a necessary prelude to modeling the flow of mobile meltwater in channel-river-moulin systems, and routing to the base and/or margins of the ice sheet.

More detailed column models of snow and firn exist, that include compaction, grain size, and other processes. Some focus on dry-snow zones, and have fine vertical resolution spanning the entire firn column with Lagrangian tracking of annual snow layers (e.g., FirnMICE: Lundin et al., J. Glac., 2017). However, they are generally too computationally expensive for ISM applications, and are not designed for ablation zones with meltwater and bare ice in summer. More general models are used in some regional climate models (RCMs) that include similar physics but with fewer layers, and are applicable both to accumulation and ablation zones.

Here we formulate a new snow-firn model, similar to those in RCMs, for use within an ice-sheet model. A relatively small number of vertical layers is used (~10), with Lagrangian tracking of explicit layers, grain size evolution, compaction, ice lenses, liquid melting, storage, percolation and runoff. Surface melting is computed from linearized net atmospheric energy fluxes, not from PDDs. The model is tested using the FirnMICE experiments, and using gridded RACMO2 modern climate input over Greenland, seeking to balance model performance with computational efficiency.

Exploring Atmospheric Drivers of Surface Melt with Weather Models

David F. Porter

Lamont-Doherty Earth Observatory, Palisades, NY

Determining the origin, timing and source of ice shelf melt will be essential to closing the surface water budget and estimating ice shelf vulnerability. To determine the total amount of water generated on glaciers and ice shelves we use a combination of weather models for fine scale processes and climate models for projections. It has recently been demonstrated that the meltwater observed around Antarctica is not captured in familiar atmospheric products such as RACMO/ANT. Although the inclusion of many crucial ice-atmosphere processes greatly improves ice sheet surface reconstructions the resolution of these continental-scale models, is simply too coarse to capture the processes important for surface melt on outlet glaciers and small ice shelves. Here we turn to higher resolution weather models with more realistic terrain crucial for capturing the strength and direction of katabatic winds, tracking of synoptic weather systems, and orographic process, all of which contribute to patterns of water input to the surface hydrologic system.

Preliminary analysis of surface melt generation in the 3.3 km resolution Antarctic Mesoscale Prediction System (AMPS) domain 3 (Powers, Manning et al. 2012) which also uses PolarWRF 3.7.1 for the dynamical core, highlights the high-degree of spatial and temporal variability over coastal Antarctica. The katabatic winds responsible for the maintenance of the Terra Nova Bay polynya accelerate down from the plateau melting the surface of the Nansen Ice Shelf along the way. A combination of enhanced sensible heating due to the strong winds and increased downward solar radiation from clear skies can result in complex patterns of melt. In the AMPS model, Reeves glacier, with some location melting approximately a third of the days in summer, has comparatively more melt than Priestly glacier to the north. Both inland locations have more melt days than the ice front. The time series of surface skin temperatures highlights the diurnal and synoptic cycles of melt. At each location, the warming above the melt threshold occurs periodically throughout the summer can be compared with observed distribution and timing of melt water. This comparison of AMPS melt and satellite observation so far reveals that melt ponds from Reeves connect earlier and appear to drain more water than the Priestly streams.

Meltwater runoff or firn refreezing? Where does the meltwater go?

Asa Rennermalm (1), Regine Hock (2), Marco Tedesco (3), Giovanni Corti (4), Federico Covi (2), Clement Miege (5), Jonathan Kingslake (3), Sasha Leidman (1)

1. Rutgers University New Brunswick, New Brunswick, NJ, United States,

2. Univ of AK-Geophysical Inst, Fairbanks, AK, United States,

3. Lamont Doherty Earth Observatory, Columbia University, Palisades, NY, United States,

4. Reed College, Portland, OR, United States,

5. University of Utah, Salt Lake City, UT, United States,

Almost half of the meltwater produced on the surface of the Greenland ice sheet infiltrate into firn layers and refreezes, while the remaining portion runs to the ocean through surface and subsurface channels. Refreezing of meltwater in firn can create impenetrable ice lenses, hence being a crucial process in the redistribution of surface runoff. Here, we present evidence of meltwater refreezing in ~20 meter deep firn cores taken from five sites along a roughly 400 meter elevation gradient in southwest Greenland in 2017. The cores are augmented with ground penetrating radar data collected at and between sites. Firn-core and radar data reveals increasing frequency and thickness of ice lenses at lower ice-sheet elevations, in agreement with previous work in the area. While this data are from the Greenland ice sheet, we discuss the implications of this work on the hydrology and surface mass balance of the Antarctic ice sheet.

Surface melt and ocean temperature control on Antarctic tidewater glacier terminus positions: implications for sea-level rise contributions from the west Antarctic Peninsula

Kiya Riverman, David Sutherland, Ryan Obermeye

University of Oregon

The west Antarctic Peninsula is one of the fastest-warming regions in the southern hemisphere, with some contribution to global sea level rise from retreating glaciers. Glacier retreat has been forced by atmospheric warming in the north west of the Peninsula and oceanic warming at depth in the south west. Along the Peninsula, we find that there is a strong difference in glacier response to warming between glaciers with grounded and floating termini, with higher mean retreat rates for glaciers with floating termini. In order to better describe future ice change along the Peninsula, we investigate the controls on glacier terminus location and flotation state along a transect of the western side of the Peninsula. We find that there is an air temperature control to floating ice extent: glaciers do not have floating tongues in areas with Mean Annual Air Temperatures (MAAT) above -6.5C. This sets a northern boundary to rapid ice retreat, and as a result, glaciers north of ~ 68.2S have remained fairly stable throughout the satellite record. Continued atmospheric warming could drive the -6.5 MAAT isotherm to the south, resulting in short-term contributions to sea level rise, but ultimately de-sensitizing the west Antarctic Peninsula glaciers to further oceanic warming by increasing the proportion of grounded (and therefore more stable) glacier termini. 

I am interested in being considered for travel support. My last degree (Ph.D.) was obtained in May of 2017. I am currently funded on a PostDoctoral Fellowship at the University of Oregon, however my fellowship only provides my salary and does not include funding for travel to meetings or meeting registration costs. In this project, I am the lone glaciologist collaborating with a group of oceanographers. I would value the input of other glaciologists on the direction my work, and this meeting could provide an excellent venue for those conversations.

Jonathan Ryan

UCLA/Brown University

The exceptional melt of the Greenland Ice Sheet since the mid-1990s has been primarily driven by rising air temperatures, increases in downward shortwave radiation, and a reduction of ice sheet albedo. Whilst our understanding of air temperatures and downward shortwave radiation are relatively well constrained, the controls on ice sheet albedo remain uncertain. Here, we develop a robust method to classify snow and ice pixels from MODIS surface reflectance products between 2000 and 2017. We use these classified maps to quantify changes in bare ice extent and investigate the role of bare ice and snow albedo to the reduction of ice sheet albedo since 2000. We find that bare ice extent has a mean extent of 232,395 km2 in the summer, equivalent to 13% of the ice sheet, and displays substantial interannual and seasonal variability. Due to the recent interest in surface hydrology in Antarctica, and the relationship between blue ice and surficial drainage networks, we believe this approach may be useful for mapping seasonal and interannual extent of blue ice, which had a similar area (234,549 km2 ) across the Antarctic Ice Sheet in the 21st century.

Precursor evidence of weakening prior to break-up of the Larsen Ice Shelves

T. Scambos, C. Shuman, and M. Klinger

Ice flux into the embayments left behind by the collapse of the Larsen A and Larsen B ice shelves surged 2- to 6-fold after their disintegration events in 1995 and 2002. Glacier imbalance in the region since the events has been persistent, with elevation changes indicating a mass loss per year of approximately twice the rate of accumulation (Scambos et al., 2014, TCryo). The proximal cause of the disintegration events was a group of processes arising from the presence of extensive surface melt lakes and hydrofracture. However, precursor changes in the ice shelves beginning more than a decade before the disintegrations have been identified, and coincide with a trend towards reduced sea ice cover and increased foehn winds.

Ice flow speeds in the Larsen A and B increased, even in the period prior to the loss of critical inboard areas of the ice shelf (which began in 1998 for the Larsen B; outboard areas were lost in 1995 along with most of the Larsen A), and elevation of the ice shelf surface decreased. Ice shelf surface lowering is interpreted as resulting from actual ice shelf thinning for this area, since field studies on both the Larsen A and B noted the upper firn of the shelf was almost completely converted to ice.

Examination of satellite images spanning 1963 - 2014 shows that Larsen B shear margins and some suture zones evolved significantly prior to significant ice shelf retreat. Overall, these changes suggest either increased ocean-driven basal melt or effects of increased surface meltwater on grounded glacier outflow can cause ‘precursor’ shelf weakening that leads eventually to disintegration. Available ocean temperature data show that modified Weddell Deep Water, having a temperature 0.1-0.4°C above the surface freezing point, is present near the former ice fronts in some 1995-2012 profiles, but to date has not been detected within the embayments near the glacier grounding lines or beneath the ice shelves in part due to limited ship access to the region.

Constraining the Spatial and Temporal Evolution of Supraglacial and Englacial Meltwater Using Radar Sounding Data

Dustin M. Schroeder, Winnie Chu, Alexander K. Kendrick, Sean Peters, Davide Castelletti

Departments of Geophysics and Electrical Engineering, Stanford University

Recent and planned airborne radar sounding surveys are providing repeat coverage of previous survey lines, enabling direct observation of temporal variation in ice sheet basal boundary conditions. In parallel, the availability and proliferation of relatively low-cost, low-power stationary sounders are producing time series observations of ice-shelf melt rates and advection rates. The development of systems with array- based and quasi-continuous-wave architectures also promises to extend these time series into 2D and 3D records. Finally, both active and passive radio sounding approaches are under development for to characterize the distribution of liquid water within planetary ice shells. Each of these radio-geophysical approached has the potential to provide observational constraint supraglacial and near-surface meltwater in Antarctica. We will discuss the challenges and opportunities for adapting these techniques (developed for subglacial, Greenland, and planetary applications) to the surface of Antarctic ice shelves.

Multi-Decadal Trends of a Supraglacial Pond Basin on the Amery Ice Shelf, East


Julian J. Spergel & Jonathan Kingslake

Lamont-Doherty Earth Observatory, Palisades, NY

The surface hydrology of Antarctic ice shelves has long-term, global impacts. Models predict a continent-wide increase in surface melting this century. Understanding how this will affect ice-shelf stability can improve predictions of global sea level. We focus on surface water storage on the Amery Ice Shelf (AIS), East Antarctica. Hundreds of meltwater ponds and channels (up to and exceeding 100 km in length) have been described in previous field accounts since the 1940’s. Using multispectral data from Landsat 4,5,7 and 8, and MODIS Terra/Aqua, and Sentinel-1 Synthetic Aperture Radar data, we track the extent of one of the largest ponds on the AIS, northwest of Clemence Massif (S72.2°, E68.7°), from 1974-2017. Since 1974, the lake, ''Big Lake,'' appears in the same location in satellite imagery several times per decade, and sometimes grows cover >7000 km2. We compare melt pond shape, growth rate and extent with temperature data from Davis Station (530 km northeast) and regional climate model output to examine how these parameters respond to climate The shape of Big Lake has become less complex since 2000, evolving from a spurred shape to a thin ovoid. Our observations lead us to hypothesize that meltwater refreezing may impart on the system a ‘memory’ of past climatic conditions, which controls lake extent. Repeated freezing of percolating meltwater decreases the permeability of the sub-surface and flattens the surrounding surface topography, resulting in more ovoid lakes. This results in expanding melt pond surface area in consecutive years with significant melt. This 'memory' maybe lost after several years of inactivity. Predicted warming could lead to increases melt pond surface area if it results in more consecutive years with significant melt. Further work is needed to test this hypothesis further.

Does ice-surface hydrology impact large-scale ice flow?

Martin Wearing

Lamont-Doherty Earth Observatory, Palisades, NY. 

It has been observed that surface-meltwater production and ponding in Antarctica is much more widespread and consistent year-to-year than first thought. The presence of this meltwater indicates a need to accurately account for its effect on ice-sheet dynamics. Much research has focused on the effect of varying quantities of water at the ice base, and the consequences for ice dynamics through basal sliding. However, less attention has been paid to the effects on ice dynamics of meltwater than is stored englacially or refreezes after entering the ice. These effects may be important in both grounded and floating sections of the ice sheet.

The refreezing of surface meltwater after it enters the ice leads to a warming of the ice in the area surrounding the refreezing site. In turn, this warming leads to a reduction in the effective viscosity of the ice, and has potential to affect the large-scale flow of the ice sheet, particularly if the production of meltwater is spatially consistent over multiple years or decades. One area where reducing ice viscosity is important is where the ice experiences large shear stresses, such as close to pinning points (islands and ice rises) and shear margins, where the resistance from lateral shear stresses is a dominant control on ice flow and contributes to large-scale stability. Therefore it is important to identify those areas that are most sensitive to englacial warming and the effect the warming would have on ice-sheet flow.

Here I assess the impact of englacial warming on the flow of an ice-shelf using a simple ice- shelf model with an idealized geometry. I impose numerous englacial-warming scenarios to compare the effects of varying spatial distribution and magnitude of englacial warming by analysing the resulting ice speed and buttressing.

Tropical Induced Surface Warming in West Antarctica

Xiaojun Yuan

Lamont-Doherty Earth Observatory, Palisades, NY. 

In the Southern Hemisphere, the strongest warming trend has occurred across the Antarctic Peninsula region from austral fall to spring and throughout West Antarctica during austral spring in the satellite era. In the same period, sea ice in the Southeast Pacific retreats significantly and upper ocean temperature increases in response to the global warming. These systematic changes are accompanied by recent thinning of the ice shelf and accelerated mass loss of the ice sheet in West Antarctica.  This work reviews recent advances in research on tropical induced warming in the Antarctic Peninsula region and West Antarctica. In addition to the El Niño-Southern Oscillation (ENSO) impacts on high latitudes, tropical to polar connections have also been discovered at the intra-seasonal timescale, associated with Madden-Julian Oscillations (MJO).  On the timescale of decades, changes in MJO phases can result in temperature and sea ice changes in the polar regions of both hemispheres. Moreover, the long-term changes in SST of the western tropical Pacific, tropical Atlantic, and North Atlantic Ocean have been linked to the rapid winter warming around the Antarctic Peninsula, while SST changes in the central tropical Pacific have been linked to the warming in West Antarctica. Rossby wave trains emanating from the tropics remain the key mechanism for tropical and polar teleconnections from intraseasonal to decadal timescales. Moreover, the SST warming in the central tropical Pacific can impact on high latitudes through altering zonal wind systems and mean meridional circulations in both hemispheres. This zonally symmetric mechanism extends tropical signals to mid-high latitudes and influences both Northern and Southern annular modes. The interaction between the Southern annular mode and high latitude responses to ENSO events also contributes to the West Antarctica warming and sea ice retreat in the recent decades.